Exploring the metabolic-immune score in advanced NSCLC treated with immunotherapy

Abstract

Identifying reliable prognostic markers in patients with advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs) is critical for clinical decision-making. This study introduces the Metabolic and Immune Score (MIS), a novel scoring system combining metabolic and inflammatory markers. A retrospective analysis was conducted on 56 patients with advanced NSCLC treated with ICIs between January 2018 and January 2024. Baseline metabolic tumor volume (MTV) and total lesion glycolysis (TLG) were calculated using ¹⁸F-FDG PET-CT imaging. Systemic inflammatory status was assessed using the Lung Immune Prognostic Index (LIPI). Median values for MTV and TLG were used as cut-off points. Primary outcomes were progression-free survival (PFS) and overall survival (OS). Patients with a LIPI score of 0 demonstrated significantly longer PFS (25.1 vs. 4.1 months, P < 0.001). High TLG levels (> 250.20) were also associated with shorter PFS (4.0 vs. 12.5 months, P = 0.021). On baseline PET-CT, median MTV and TLG were determined as 57.29 cm³ and 250.20, respectively. The MIS, derived from the combination of these two parameters, stratified patients into good, intermediate, and poor prognostic groups. Significant differences in PFS were observed among MIS groups (25.1, 6.3, and 1.5 months; P < 0.001), whereas median OS was not yet reached in the favorable group, and was 13.1 and 5.0 months in the intermediate and poor groups, respectively (P = 0.029). The MIS combines LIPI and TLG, providing a useful tool for predicting outcomes in advanced NSCLC patients treated with immunotherapy. Further validation in larger cohorts is warranted.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-16788-7.

Introduction

Non-small cell lung cancers (NSCLC) represent approximately 80% of all lung cancer cases, which remain the leading cause of cancer-related mortality globally^1,2. The advent of immune checkpoint inhibitors (ICIs) has significantly improved survival in metastatic NSCLC. However, responses are highly variable, and only a subset of patients achieves meaningful clinical benefit^3–5.

To predict response and prognosis with ICIs therapy, researchers have extensively investigated various biomarkers related to the tumor microenvironment and tumor biology, with PD-L1 level being one of the most studied indicators⁶. While PD-L1 expression has been extensively studied, it remains an imperfect predictive biomarker, as responses can be observed in patients with low or absent PD-L1 levels, and conversely, some high PD-L1-expressing patients do not respond^7,8.

Numerous biomarkers have been investigated for their potential to predict immunotherapy response in NSCLC. These include genetic markers such as tumor mutational burden (TMB) and microsatellite instability (MSI); immune-based biomarkers including CD8 + T cell infiltration, regulatory T cells (Tregs), tertiary lymphoid structures (TLS), myeloid-derived suppressor cells (MDSCs), and Immunoscore^9–17. Also, radiomics-based features on baseline and serial CT imagings, have shown promising predictive performance in recent studies^18–20. Additionally, peripheral blood cell subpopulations, such as the neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) have shown potential as cost-effective, non-invasive predictors^21,22. Despite these advances, none of these biomarkers have yet achieved widespread clinical validation or consensus as definitive predictive tools. Consequently, there is a strong need to explore biomarkers that may more accurately reflect tumor characteristics.

The standard imaging modality for staging in NSCLC includes 18-fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET-CT)^23,24. Beyond staging, quantitative metrics derived from ¹⁸F-FDG PET-CT, such as metabolic tumor volüme (MTV), total lesion glycolysis (TLG), and standardized uptake value (SUV), have demonstrated promising prognostic value in NSCLC patients receiving ICIs therapy^25–27.

In parallel, Lung Immune Prognostic Index (LIPI) has emerged as a promising predictive tool based on derived neutrophil-to-lymphocyte ratio(dNLR) and plasma lactate dehydrogenase (LDH) levels. LIPI stratifies patients into three prognostic groups—good, intermediate, and poor—based on specific dNLR and LDH cutoffs. This index has demonstrated prognostic value across multiple studies for ICI-treated patients^28–30 .

Since metabolic tumor burden and systemic inflammation each capture different but complementary aspects of cancer biology—one reflecting how aggressive the tumor is, the other indicating the host’s immune response—their combination into a single score is expected to offer a more complete and accurate prognostic picture than either marker alone. Although several studies have explored the combination of immune and metabolic parameters, many were limited by methodological heterogeneity, lack of validation, or complexity that restricted clinical applicability.

Although both LIPI and ¹⁸F-FDG PET-CT–derived quantitative metrics have shown individual prognostic relevance in advanced NSCLC, their combined clinical utility remains underexplored²⁵. Given that LIPI reflects host-related systemic inflammation and PET-based parameters represent tumor burden and metabolic activity, an integrated model may offer a more holistic prognostic stratification. To address this unmet need, we developed the metabolic and immune score MIS and aimed to evaluate its ability to predict outcomes in patients receiving ICIs.

Material & method

This study was conducted in accordance with the “Declaration of Helsinki” and approved by the local ethical committee of Ankara University Faculty of Medicine (Application and approval dated 20.05.2021, with the registration number 2121/115, is available). Due to the retrospective nature of the study, Ankara University Faculty of Medicine Ethics Committee waived the need of obtaining informed consent.

Patients

Patients with histologically confirmed metastatic NSCLC, either de novo or recurrent, who received ICI therapy between January 2018 and January 2024, were screened. ICI treatment was given as monotherapy or in combination with chemotherapy, in either first-line or later treatment settings.

Patients were eligible if they had a baseline ¹⁸F-FDG PET-CT performed within four weeks before starting ICI therapy and had available laboratory data to calculate the LIPI.

Patients with small cell histology, missing pre-treatment PET-CT imaging, or incomplete blood tests required for LIPI, were excluded.

Demographic and clinical data (age, sex, ECOG performance status, smoking history, comorbidities, and stage at initial diagnosis), PD-L1 status (if available), prior and concurrent treatments (if any) before ICI therapy, and laboratory values within the four-week pre-ICI period, including hemogram and LDH levels were collected retrospectively. Tumor PD-L1 expression was evaluated locally using immunohistochemical staining on pretreatment and archival biopsy samples.

LIPI score calculation

Laboratory evaluations, including complete blood cell counts and LDH levels, were performed within four weeks prior to treatment initiation. The normal LDH range was defined as 135 to 225 units per liter (U/L). According to the LIPI score classification, a dNLR (neutrophils/[leukocytes − neutrophils]) greater than 3 and an LDH level above the upper limit of normal are considered unfavorable prognostic factors. Using these criteria, participants were stratified into three prognostic subgroups: good prognosis (LIPI = 0), intermediate prognosis (LIPI = 1), and poor prognosis (LIPI = 2)²⁸ .

¹⁸F-FDG PET-CT imaging protocol and data analysis

¹⁸F-FDG PET-CT was performed within four weeks prior initiation of ICI therapy. Patients fasted for at least 6 h, with blood glucose levels confirmed to be below 200 mg/dL before PET-CT imaging. Patients were injected with ¹⁸F-FDG following current guidelines (median activity 255 MBq, range 106–446 MBq, median 3.5 MBq/kg), and whole body PET-CT acquisition was started after a median interval of 60 min.

The imaging procedure included a non-contrast CT scan (120 kV, 70 mA, tube rotation time of 0.5s per rotation, a pitch of 1.375, and a slice thickness of 3.3 mm) followed by a PET scan of 2 min/bed position, covering the area from the vertex to the proximal femur.

Two experienced nuclear medicine specialists, who were blinded to clinical outcomes, have retrospectively reviewed PET-CT images on PET VCAR Software (GE Healthcare, Waukesha, WI, USA). Each reviewer assessed the images independently, and any discrepancies were resolved by consensus to ensure interobserver consistency. Semiautomated volume of interests (VOI) were drawn around target lesions. A threshold of 41% of SUVmax was used for lesion delineation, in line with prior studies that applied this method^31–33. All hypermetabolic lesions (both primary and metastatic) visually recognized from the surrounding physiological uptake were taken into account. In patients with widespread metastatic disease, only 20 lesions with highest uptake were accepted. Physiological uptake sites and lesions of high uptake related with benign/inflammatory process were excluded. SUV calculations included SUVmax and SUVpeak of the most intense lesion, SUVmean for each hypermetabolic lesion, total MTV as the sum of all lesions, and TLG, calculated by summing SUVmean multiplied by the metabolic volume of each lesion. Since no standardized cut-off values for MTV and TLG exist, median values were used as cut-offs. According to baseline PET-CT evaluations, the median MTV was 57.29 cm³, and the median TLG was 250.20.

Outcomes

The primary outcome was PFS, defined as the time from the initiation of ICI treatment to the date of documented disease progression or death from any cause, whichever occurred first. The secondary outcome was OS, defined as the time from the initiation of ICI treatment to death from any cause, or censored at the date of last follow-up for patients still alive. Disease progression was evaluated based on iRECIST criteria³⁴. The data cut-off date was October 25, 2024. Patients who were alive or progression-free at their last follow-up were censored at that date.

Statistical analysis

SPSS 25.0 for Windows was used for statistical analyses. Descriptive statistics were presented as mean ± standard deviation for normally distributed variables, median (IQR) for non-normally distributed variables, and frequency (%) for nominal variables. The significance of differences in values between groups was tested using the t-test or Mann-Whitney U test. Nominal variables were analyzed using the chi-square test or Fisher’s exact test. Kaplan-Meier analysis was used to estimate survival. Potential factors affecting OS were assessed in univariate analysis, and the ones with P < 0.05 were included in the multivariate analysis. The Cox proportional hazard models were used to assess the hazard ratios (HR) and two-sided 95% CIs. Receiver operating characteristic (ROC) analysis was performed to evaluate the prognostic performance of the MIS score. The area under the curve (AUC) was calculated to assess the discriminatory ability of MIS. Values with 2-sided P < 0.05 were considered statistically significant.

Results

Baseline characteristics

A total of 56 metastatic NSCLC patients, with a median age of 61 years (IQR: 33–76) and 78.6% male, were included. At diagnosis, 20 patients (35.7%) were metastatic. At diagnosis, 16% of patients were stage II, 48% stage III, and 36% stage IV. Histology was non-squamous in 67.9% and squamous in 32.1%. PD-L1 expression was assed in 48 patients, with 69.5% testing positive. In 52 patients (92.9%), ICI was administered as second-line or later therapy, and 14.3% received ICI in combination with chemotherapy (Table 1).

Table 1.

Clinical characteristics at immunotherapy Initiation.

Characteristic	(n = 56)
Age, years, median (IQR)	61 (33–76)
Sex, n (%)
Male	44 (78.6)
Female	12 (21.4)
Comorbidities, n (%)
Hypertension	20 (35.7)
Diabetes	6 (10.7)
CVD	8 (14.3)
COPD	10 (17.9)
Smoking history, n (%)
Never	18 (32.1)
Ex-smoker	30 (53.6)
Current smoker	8 (14.3)
ECOG performance status, n (%)
0–1	49 (87.5)
≥2	7 (12.5)
Metastatic presentation type, n (%)
Post-treatment metastatic	36 (64.3)
De novo metastatic	20 (35.7)
Site of metastasis, n (%)
Liver	10 (17.9)
Lung	26 (46.4)
Bone	21 (37.5)
CNS	14 (25)
PD-L1 expression, n (%)
<1	9 (18.8)
1–49	22 (45.8)
≥50	17 (35.4)
Previous chemotherapy, n (%)	49 (87.5)
Previous radiotherapy, n (%)
No RT	27 (48.2)
Definitive RT	22 (39.2)
Palliative RT	7 (12.5)
ICI agent, n (%)
Nivolumab	43 (76.8)
Pembrolizumab	13 (23.2)
ICI line, n (%)
First line	4 (7.1)
Second line or later line	52 (92.9)
ICI combined with chemotherapy, n (%)	8 (14.3)

Abbreviations: CNS: Central nervous system, COPD: Chronic obstructive pulmonary disease, CVD: Cardiovascular disease, ECOG: Eastern Cooperative Oncology Group, ICI: Immune checkpoint inhibitor, IQR: Interquartile range, PD-L1: Programmed death ligand-1. Note: PD-L1 level was available in 48 out of 56 patients. Percentages were calculated based on the evaluable cases.

Baseline PET-CT parameters

On baseline PET-CT images, patients had a median lesion count of 4 (IQR: 1–18), a median MTV of 57.29 cm³ (IQR:3.68-21083.44), and a median TLG of 250.20 (IQR:9.10-6340.40). Other baseline PET-CT parameters of the patients are presented in Table 2. Median values were used as cut-offs for MTV and TLG. Patients with MTV ≤ 57.29 cm³ were classified as low-MTV, and those with MTV > 57.29 cm³ as high-MTV; similarly, patients with TLG ≤ 250.20 were classified as low-TLG, and those with TLG > 250.20 as high-TLG. Accordingly, there were 28 patients with low-MTV and 28 patients with high-MTV, as well as 28 patients with low-TLG and 28 patients with high-TLG group.

Table 2.

¹⁸F-FDG-PET imaging parameters.

Parameter	Value
Measurable lesions, number	4 (1–8)
Metabolic tumor volume, cm3, median (IQR)	57.29 (3.68–21083.44)
Total lesion glycolysis, median (IQR)	250.20 (9.10–6340.40)
SUVmean, median (IQR)	8.03 (2.08–28.06)
SUVpeak, median (IQR)	10.98 (2.03–38.80)
SUVmax, median (IQR)	15.14 (2.66–48.12)

Abbreviations: IQR, Interquartile range; SUV, Standardized uptake value; TLG, total lesion glycolysis; ¹⁸F-FDG, fluorine-18 fluorodeoxyglucose.

Survival outcomes

Median follow-up time was 11.4 months (IQR: 1.2–74.2). Progression was observed in 35 patients (62.5%), with a median PFS of 6.5 months (95% CI: 3.6–9.4).

The LIPI was 0 in 21 patients, 1 in 19 patients, and 2 in 16 patients. Patients with a LIPI of 0 had significantly longer PFS compared to those with LIPI > 0 [25.1 months (95% CI: 9.0–41.1) vs. 4.1 months (95% CI: 1.1–7.2), P < 0.001]. While MTV assessed by PET-CT showed no significant impact on PFS [6.9 months (95% CI: 0.7–14.3) vs. 6.5 months (95% CI: 2.5–10.5), P = 0.369], patients in the TLG-high group had significantly worse PFS [4.0 months (95% CI: 0.2–7.7) vs. 12.5 months (95% CI: 2.1–22.9), P = 0.021] (Fig. 1a, b and c).

Fig. 1.

Kaplan-Meier survival analyses of: Progression-free survival by a) LIPI, b) Total metabolic tumor volume (MTV), c) Total lesion glycolysis (TLG). Overall survival by d) LIPI, e) Total metabolic tumor volüme (MTV), f) Total lesion glycolysis (TLG).

In multivariate analysis, liver metastasis [HR: 2.7 (95% CI: 1.2–6.3), P = 0.016], LIPI > 0 [HR: 5.6 (95% CI: 2.1–14.7), P < 0.001], and high-TLG [HR: 4.1 (95% CI: 1.2–14.9), P = 0.029] were identified as independent factors associated with worse PFS (Supplementary Table 1).

A total of 27 patients (48.2%) died, and the median OS was 14.4 months (95% CI: 3.4–25.4). Patients with a LIPI score of 0 had better OS compared to those with LIPI > 0 [NR vs. 8.9 months (95% CI: 5.0–12.8), P = 0.001]. In the MTV-high group, OS was shorter but did not reach statistical significance [NR vs. 11.8 months (95% CI: 4.9–18.7), P = 0.089]. OS was significantly shorter in the TLG-high group [8.5 months (95% CI: 3.6–13.5) vs. NA, P = 0.004] (Fig. 1d, e and f).

In multivariate analysis, LIPI > 0 [HR: 3.59 (95% CI:1.3–9.7), P = 0.011] and high-TLG [HR: 2.31 (95% CI:0.9–5.5), P = 0.048] were identified as independent factors associated with worse OS (Supplementary Table 2). In ROC analysis for predicting overall survival, TLG demonstrated a moderate prognostic ability with an AUC of 0.693 (95% CI: 0.5–0.8, p = 0.013), whereas the LIPI score had an AUC of 0.634 (95% CI: 0.5–0.8, p = 0.085).

We developed a scoring system, called ‘Metabolic and Immune Score (MIS)’, that combines the LIPI and TLG, both of which were shown to be independent prognostic factors for survival. Based on their scores, patients were categorized into good (0 points), intermediate (1–2 points), and poor (3 points) prognosis groups (Table 3). Consequently, 14 patients were classified in the favorable prognosis group, while 31 were in the intermediate group, and 11 in the poor prognosis group. Among these groups, the median PFS was 25.1 months (95% CI: 19.8–30.3), 6.3 months (95% CI: 4.8–7.8), and 1.5 months (95% CI: 0.5–2.5), respectively (Fig. 2a), with significant differences in PFS between the groups (P = 0.011 for good vs. intermediate, P < 0.001 for good vs. poor, and P = 0.001 for intermediate vs. poor). Similarly, the median OS was not reached in the good prognosis group, while it was 13.1 months (95% CI: 7.2–18.9) in the intermediate and 5.0 months (95% CI: 1.7–8.3) in the poor prognosis groups, with statistically significant differences in OS between the groups (P = 0.029, P = 0.001, and P = 0.016, respectively) (Fig. 2b).

Table 3.

Definition of the MIS and corresponding prognostic Groups.

Varıable	Value	Score
LIPI	0	0
	1	1
	2	2
TLG	low	0
	high	1
PROGNOSTIC GROUP	TOTAL MIS SCORE
Good	0
IntermEdIate	1–2
Poor	3

Abbreviations: LIPI, Lung Immune Prognostic Index; TLG, Total Lesion Glycolysis; MIS, Metabolic and Immune Score. Note: MIS was defined as the sum of LIPI and TLG scores (0–3). TLG was classified as “low” or “high” based on the cohort’s median value (250.2).

Fig. 2.

Kaplan-Meier survival analyses of a) Progression-free survival, b) Overall survival by MIS groups.

In the ROC curve analysis, the MIS score showed an AUC of 0.631 (95% CI: 0.5–0.8), with a P-value of 0.104 (Fig. 3).

Fig. 3.

Roc curve analysis of Metabolic and Immune Score (MIS).

After excluding eight patients who received a combination of ICI and chemotherapy, the MIS score remained a useful prognostic tool in the 48 patients who received immunotherapy alone. OS was not reached in 11 patients within the good prognostic group, while it was 11.8 months (95% CI: 8.070–15.651) in 27 patients classified as intermediate and 2.7 months (95% CI: 0.1–5.8) in 10 patients in the poor prognostic group. The differences in OS were statistically significant (P = 0.014 for good vs. intermediate; P < 0.001 for good vs. poor; P = 0.004 for intermediate vs. poor).

Discussion

In this study, we developed and evaluated the Metabolic and Immune Score (MIS), a combined model integrating PET-CT–derived metabolic tumor activity (TLG) and systemic inflammation (LIPI), to predict prognosis in patients with metastatic non-small cell lung cancer treated with immune checkpoint inhibitors. MIS effectively stratified patients into distinct prognostic groups, with significant differences in both progression-free and overall survival. Notably, those in the favorable-risk group had a median PFS of 25.1 months, compared to only 1.5 months in the poor-risk group.

Studies showed that higher MTV values correlated with an increased tumor burden^25–27. In our study, MTV did not demonstrate a significant impact on survival, suggesting that tumor volume alone may not sufficiently reflect the biological behavior of the tumor. Active proliferative cancer cells rapidly absorb glucose and produce lactic acid, impairing T cell function by reducing cytokine production, which negatively affects cytotoxic T cells. This indicates that TLG, which reflects both metabolic activity and tumor volume, may show more aggressive tumor behavior compared to MTV. The absence of a significant association between MTV and survival outcomes may emphasize that tumor size alone is insufficient for prognosis, whereas high TLG values may indicate increased tumor proliferation and a challenging tumor microenvironment. Our findings did not show a significant association between MTV and survival, which contrasts with previous studies, including the meta-analysis by Zhu et al.³⁵. Our relatively small sample size could have limited our ability to detect more subtle associations. While we used median values to define MTV thresholds, Zhu et al. highlighted that clinically meaningful cut-offs often lie between 50 and 100 cm³. Differences in image segmentation approaches and the heterogeneity of patient populations across studies may also contribute to the variability in MTV’s prognostic performance. Consequently, TLG may serve as a more sensitive biomarker for predicting response and prognosis in mNSCLC patients treated with ICIs.

Our ROC analysis revealed that the MIS score had a modest discriminatory accuracy, with an area under the curve (AUC) of 0.631. MIS successfully divided patients into distinct risk groups, indicating its potential use in cohort-level prognostic evaluation, even though this value represents limited predictive power at the individual-patient level. Given that it is a dual-parameter, simplified model that incorporates baseline metabolic and inflammatory markers, this is not surprising. To preserve conceptual simplicity and clinical usability, we chose to give both variables equal weight in the MIS score, even though TLG showed statistically significant prognostic power while LIPI did not. Together with TLG, LIPI is still a widely accepted and biologically plausible indicator of systemic inflammation, which allows the score to account for both host immune status and tumor-intrinsic metabolic activity. This two-pronged approach is consistent with growing evidence that host and tumor factors affect immunotherapy response. However, there are a number of chances to improve its prognostic precision. Integrating additional established biomarkers such as PD-L1 expression or tumor mutational burden (TMB), which remain a primary clinical stratification tool, could provide valuable complementary information^4,36. Moreover, radiomic analysis of PET/CT images has emerged as a promising avenue to extract intratumoral heterogeneity and microenvironmental features that go beyond conventional metabolic parameters³⁷. Similarly, peripheral blood markers such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and monocyte-to-lymphocyte ratio (MLR) have been associated with immune status and treatment outcomes, and may refine prognostic modeling when combined with MIS³⁸. Finally, the incorporation of dynamic, longitudinal data—such as serial changes in PET parameters or inflammatory markers during therapy—may allow MIS to evolve into a real-time monitoring tool. Additionally, in ROC analysis, TLG alone demonstrated a higher AUC (0.693) than LIPI (0.634), and was the only parameter to reach statistical significance. Future prospective studies should explore these extensions to build upon the current model and better capture the multifaceted biology of immunotherapy response.

In addition to metabolic and inflammatory markers, we also assessed other clinical variables. Although de novo metastatic disease showed a trend toward poorer survival in univariate analysis, it did not remain significant in multivariate models. This suggests that its apparent prognostic impact may be explained by more dominant biological factors such as tumor burden or systemic inflammation.

A significant limitation of our study is its retrospective nature and the inclusion of a limited patient population from a single center, which restricts the generalizability of the findings. The follow-up period of approximately 11.4 months is relatively short, which may not adequately capture the natural progression of the disease and treatment response. Also, the inclusion of a heterogeneous patient group may obscure the influence of various treatment protocols and disease stages on the outcomes. To deal with this, subgroup analyses were done based on treatment line and modality. While there were no statistically significant differences, these results are provided as supplementary data. After all, as this is the first study in which the MIS score has been proposed and applied, our findings should be interpreted as exploratory. These limitations highlight the need for further validation of the findings in larger, multi-center populations.

Despite these limitations, this study has notable strengths. It highlights the prognostic value of metabolic and immune parameters in metastatic NSCLC patients receiving immunotherapy. By assessing both TLG and the LIPI, it offers a comprehensive view of tumor biology and immune status. The use of PET-CT for evaluating tumor burden enhances the accuracy of prognostic predictions. Furthermore, the development of the MIS presents a novel approach to predicting PFS and OS, contributing valuable insights to the field of cancer treatment.

In conclusion, our findings underscore the importance of comprehensive assessments that incorporate both metabolic and immune factors in guiding treatment decisions in metastatic NSCLC patients receiving ICI treatment. The developed MIS may serve as a practical tool for clinicians, allowing for better risk stratification and personalized treatment approaches. These findings warrant further prospective validation before clinical implementation.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by BBK, ED, İM, and EBK. The first draft of the manuscript was written by BBK, ED, and İM. MSA contributed to data collection and the provision of study materials. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was conducted in accordance with the “Declaration of Helsinki” and approved by the local ethical committee of Ankara University Faculty of Medicine (Application and approval dated 20.05.2021, with the registration number 2121/115, is available). Due to the retrospective nature of the study, Ankara University Faculty of Medicine Ethics Committee waived the need of obtaining informed consent.

Consent for publication

Not applicable. This study does not contain any individual person’s data in any form.